There is a lack of confidence in present in vitro disease models and drug efficacy tests, as they do not properly recapitulate the dynamic physiology and pathophysiology of the human organism. This challenge is particularly acute in oncology: present tools to study drug responses fail to faithfully mimic the patient’s tumor microenvironment (TME) and thus have not kept up with tumor biology and drug testing needs. As a measure of this problem, on average less than 4% of oncology drugs in clinical trials end up being FDA-approved, a dismal approval rate that has dire social repercussions such as high cancer drug prices and difficult accessibility. We have developed a suite of microfluidic platforms that address this problem by multiplexing the delivery of drugs to intact-TME human biopsies, altogether bypassing animal testing. We have developed and patented a microdissection methodology that allows for producing large numbers of regularly-sized, cuboidal micro-tissues (“cuboids”) from a single tumor biopsy. We have developed arrays of microfluidic traps for cuboid perfusion and a very high-throughput automated robotic placement of mouse and human cuboids in 384-well plates. With these approaches, it will soon be possible to bypass animal testing and perform direct testing of drugs using only human tumors. Since these new-generation tests preserve the TME intact, we envision that they will minimize FDA failure rates and will contribute to alleviate the cost of cancer drugs. In this talk, I will also cover innovative 3D printing approaches of general applicability to the fabrication of complex biomicrofluidic systems such as organs-on-chips.